Intel Foundry Faces Another Major Setback: Reports Indicate Intel 18A Process Yield at Only 10%, Unable to Enter Mass Production

Previously, Intel aimed for industry leadership by 2025, announcing that the Intel 18A (1.8nm) CPUs, Panther Lake and Clearwater Forest, were underway and showing promising health.

However, recent reports reveal a shocking development: Intel's 18A process, considered the "turning point" for Intel Foundry Services (IFS), is reportedly achieving a yield of just 10%, far below what is required for mass production.

With performance across all divisions falling short of expectations, Intel is struggling to achieve a breakthrough in the industry. This is especially true for Intel Foundry Services, a "core strategy" in the company’s economic recovery plan, which is now facing significant market adoption challenges due to fierce competition and underwhelming results. According to reports, Intel's much-anticipated 18A process cannot reach the yields necessary for high-volume production, compounding its challenges.

Although these reports are difficult to believe, they may reflect the "harsh reality" given Intel’s recent struggles. Broadcom, allegedly one of Intel’s "key" foundry clients, is reportedly dissatisfied with the Intel 18A process. Broadcom engineers claim that the process is unsuitable for mass production, primarily due to its low yield. It is said that Broadcom has canceled its orders with Intel and is now seeking alternative solutions.

The poor yield and underperformance of Intel Foundry Services are reportedly among the primary reasons for the dismissal of Intel CEO Pat Gelsinger. Additionally, Intel’s failure to secure larger subsidies from the U.S. government was the final straw leading to his termination. With the outlook for Intel Foundry Services increasingly bleak, rumors of a potential sale of the division are intensifying.

While Intel struggles at this node, competitor TSMC appears to be satisfied with its own progress. Despite being "slightly behind" in node size, TSMC’s N2 (2nm) process reportedly outperforms Intel 18A due to significantly higher SRAM density. This is a crucial factor, as SRAM density plays a key role in determining node efficiency and performance, and it appears TSMC has gained the upper hand once again.

With key "Intel Foundry advocates" departing, Intel may shift its focus back to its manufacturing and product divisions. This raises the possibility that a sale or merger of Intel Foundry Services is still on the table.